我在一個月前預測八月將是修正月的看法被證明是錯誤的，因為2025年八月成為三大美股指數連續第四個正回報的月份。雖然八月的月度回報較前面三個月小，但道瓊斯指數、標普500和納斯達克指數仍分別上漲約3%、近2%和1.6%。除了包括多數AI科技股在內的穩健企業收益和一些稍微有利的宏觀數據外，八月得以受美國聯儲局鮑威爾在傑克森霍爾會議中意外偏鴿派言論點燃降息希望的影響而有所支撐。
儘管美股市場展現驚人韌性，我仍再次預測九月（傳統上是全年最弱的月份）將會有修正。不要誤會，我對美國市場的中長期展望仍保持樂觀，但短期內仍需技術性修正，方能繼續上升趨勢。此外，這次修正很可能是相對短暫且溫和的回調，幅度約4至5%，持續一到兩周，而非長期且深度調整。經歷過過去20.5周約30%的反彈漲勢後，市場非常需要技術性修正，但利好基本面及技術面背景將支持市場下行空間有限，避免出現10%或更大的調整。
九月可能較弱的主因
作為第三季度的最後一個月，九月因收益報告減少，對市場情緒幫助有限，雖然AI第二大硬體基礎設施股、美國市值1.4兆美元的博通公司將於九月四日公佈業績，這或將在英偉達符合預期的結果，以及Marvell Tech上周疲弱的表現之後，提振AI科技股情緒，但整體九月仍難期待太多公司利好消息支撐。
在宏觀數據方面，九月經濟數據點數量與前月相當，故不致成為不利因素。經過上月部分消費者支出價格指數（PCE）符合預期及其他高於預測的通脹數據後，市場將關注宏觀成長及就業相關報告，尋找利率前景線索，包括九月五日的非農就業及失業數據。不過，市場已高度預期九月降息，宏觀數據面風險偏向利空多於利多。簡言之，即使宏觀數據或聯儲局官員言論偏正面，也難大幅提振情緒，反倒較差的成長或就業數據可能令市場失望。
在貿易及關稅政策方面，上月已有不少利多消息，九月恐難復現。當然，中美關稅協議仍有潛在正面可能，但我擔心正式協議恐須到九月後，因為最後期限已延後至十一月八日。在此期間，雙方在一對一談判中可能仍會有較強硬言辭。
最後，雖然我通常不鼓勵依賴「季節性因素」，但九月的長期歷史表現仍引人擔憂。更重要的是，標普500剛連續第四個月錄得正報酬，並累積近30%、長達20.5周的大幅牛市，且未曾出現3%的修正。從技術面看，市場需要展開技術性調整以緩解超買壓力，方能延續長期上升趨勢。
總結來說，我預期九月將是修正月，出現短暫2至3周的輕度至中度回調，隨後市場將在2025年最後一季或2026年繼續其三年多牛市。投資者應持有約20至40%的戰術現金作為回調時的買入準備。
September could be the correction month for US stock after August fails to pullback
PoMy prediction a month ago that August would be the month of correction proves to be wrong as August 2025 turns out to the 4th positive month for all three major US stock indexes. Although the calendar month return is smaller than the previous 3 months, Dow Jones, S&P500, & NASDAQ are still up 3%, almost 2%, & 1.6% respectively for the month of August. Besides solid corporate earnings including most AI tech stocks & some mildly favourable macro data, August is probably saved by the surprisingly dovish Jackson Hole’s comment by US Fed Powell to reignite rate-cut hope.
Despite this amazingly resilience of US stock market, I once again forecast that it would correct this month – traditionally weakest calendar month of the year – September. Do not get me wrong, I am still optimistic about US market’s mid & long-term outlook but still see a need of technical correction in the short term before it can continue its upward trend. Moreover, this correction could be a relatively short & mild pullback of around 4-5% for a fortnight or slightly longer instead of a long & deep correction. After a roughly 30% rebound rally to record level in the past 20.5 weeks, the market is in big need of a technical correction but the favourable fundamental & technical background would underpin the market’s downside to avoid a bigger 10% or more adjustment.
Key factors ‘September’ could be a weaker month this year
As the last month of 3Q, September has much less earnings report to help market sentiment although AI’s second biggest hardware infrastructure stock – the US$1.4 trillion Broadcom Inc. will report on September 4th that may help boost AI tech sentiment after the in-line result of Nvidia & the weak earnings of Marvell Technology last week. Still, September may not get too much support from company’s event.
On the macro data side, September has the same number of economic data point and hence that should not be an unfavourable factor for it. After seeing some in-line PCE & other higher-than-forecast inflation number last month, the market will focus on macro growth- & job-related report for hint of interest rate outlook including the September 5th Non-Farm Payroll & Unemployment number. However, as the market is already pricing in a high chance of a September rate-cut, macro data may have more downside risk than upside. In short, even favourable macro number or Fed officials’ comment may not be able to help sentiment much but less-good growth or job data could disappoint the market.
